On 2 June 2014 18:42, Kalle Valo <kvalo@xxxxxxxxxxxxxxxx> wrote:
> Bartosz Markowski <bartosz.markowski@xxxxxxxxx> writes:
>
>> Firmware 10.x supports up to 8 virtual AP interfaces
>> (comparing to 7 for main firmware). Previous vdev map
>> initialization was missing enough space for 8 + 1 vdevs
>> (we may spent one for mac monitor), due to wrong define used.
>>
>> Use correct one - TARGET_10X_NUM_VDEVS - for 10.x firmware.
>>
>> Signed-off-by: Bartosz Markowski <bartosz.markowski@xxxxxxxxx>
>
> So what is the actual bug you are fixing? Previously with 10.x it was
> possible to get only 7 VIFs, even though we advertised 8 to user space,
> and with your fix we get the full 8 VIFs?
For CAC, we use one VDEV to start monitor interface. In case of 10.X
firmware we advertise support up to 8 VAPs, but if we spent one for
monitor interface, only 7 left. I've noticed we fail on .add_interface
when trying to add 8th AP, here:
bit = ffs(ar->free_vdev_map);
if (bit == 0) {
ret = -EBUSY;
goto err;
}
and this lead me to initialization code for vdev_map
ar->free_vdev_map = (1 << TARGET_NUM_VDEVS) - 1;
We have an API split for main and 10.x firmware (incl. number of
vdevs, target fw is able to handle), but here we missed this split.
Ben has a valid point, the TARGET_10X_NUM_VDEVS claims to be 16, so
there's an inconsistency between what we adverts to mac in max
interfaces, but I'm not sure if this is such a big deal.
> It would be good to clear have that in the commit log so that anyone can
> understand what bug is fixed.
Do you want me to send a v2 with just an updated commit (better user
impact description)? (No patch content changes)
-Bartosz
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
